Discussion

This case represents a rare visceral variant of Lemierre's syndrome in which the intra-abdominal venous system served as the substrate for septic thrombophlebitis, mirroring the pathophysiology of classic jugular vein involvement. F. necrophorum, an obligate anaerobic gram-negative rod, elaborates potent endotoxins, hemolysins, and platelet-aggregating factors capable of inducing locoregional coagulation and metastatic embolic seeding [5]. The temporal and anatomic relationship between the hepatic abscess, the adjacent hepatic vein thrombophlebitis, and the bilateral peripheral wedge-shaped pulmonary nodules is most consistent with the liver-derived focus acting as the embolic source of the septic pulmonary emboli; however, this remains a clinical and radiologic inference rather than a directly proven mechanism, and an alternative unrecognized primary source cannot be entirely excluded.

Nomenclature and diagnostic reasoning merit explicit discussion, since the absence of pharyngitis in this patient could reasonably raise the question of whether the label "Lemierre's syndrome" is appropriate at all. We acknowledge that this patient never had sore throat, and we do not propose an oropharyngeal origin. Rather, we apply the term in its now widely used variant sense: a primary anaerobic focus, contiguous septic thrombophlebitis of the draining vein, and metastatic septic embolization arising from that infected thrombus. In this patient all three elements were present, with the hepatic vein substituting for the internal jugular vein. An alternative and entirely reasonable mechanistic sequence, and one we considered, is that anorectal mucosal inoculation during receptive anal intercourse produced a clinically silent proctocolitis, with portal venous translocation of Fusobacterium seeding the liver, formation of the hepatic abscess, secondary hepatic vein thrombophlebitis, and finally embolization through the inferior vena cava and right heart to the pulmonary arteries. We emphasize that this proposed pathway does not conflict with the diagnosis; it is precisely the mechanism that defines the visceral variant, in which the splanchnic venous system replaces the cervical venous system as the route of dissemination. Whether the portal of entry was anorectal, another gastrointestinal site, or occult, the syndromic construct and its clinical implications remain the same.

We nonetheless sought evidence for an anorectal or gastrointestinal portal of entry and did not find it. The patient denied anorectal pain, tenesmus, discharge, bleeding, diarrhea, and abdominal pain; perianal and digital rectal examination were unremarkable; and contrast-enhanced CT of the abdomen and pelvis showed no rectal or colonic wall thickening, pericolic stranding, perirectal collection, portal or mesenteric venous thrombus, or periportal gas. Sexually transmitted infection screening, including HIV and syphilis serology, was negative. A genitourinary source was likewise considered, with no imaging evidence of nephritis or prostatitis. Endoscopic and anoscopic evaluation were not performed, and this is a limitation: a subclinical proctitis cannot be formally excluded, and nucleic acid amplification testing of a rectal swab for Neisseria gonorrhoeae and Chlamydia trachomatis was not obtained. In a patient with comparable epidemiologic risk, we would now recommend rectal nucleic acid amplification testing and a low threshold for anoscopy, even in the absence of anorectal symptoms, since establishing the portal of entry has both prognostic and public health value. The definitive portal of entry in this case therefore remains unidentified, which is not unusual in the visceral variant of Lemierre's syndrome, where the primary focus is reported as occult in a substantial minority of cases.

The severe thrombocytopenia in this critically ill patient prompted evaluation for a hematologic process, and bone marrow biopsy was undertaken to exclude primary marrow pathology; in retrospect, the rapid and sustained platelet recovery that paralleled clinical improvement was itself strongly suggestive of a consumptive, sepsis-related etiology, and a period of watchful waiting with serial platelet counts may have obviated the need for biopsy in a similar future presentation. A single positive lupus anticoagulant (dRVVT normalized ratio 1.29; reference < 1.21) was identified during acute illness; infection-associated lupus anticoagulant is a recognized phenomenon, and endothelial disruption with transient antiphospholipid antibody generation has been described during acute Fusobacterium infection [6], but transience has not been demonstrated in this patient, since only one measurement is available. We therefore describe this as a suspected infection-associated lupus anticoagulant rather than a confirmed transient phenomenon. Repeat serologic testing at three months, intended to confirm resolution, had not yet been performed at the time of writing.

Hepatic vein thrombophlebitis was a key feature of this case, and the decision regarding anticoagulation warrants explicit discussion. Anticoagulation was not administered. This decision reflected the infectious etiology of the thrombophlebitis, the expectation that antimicrobial therapy would address the primary driver of thrombus formation, and the bleeding risk posed by a platelet count of 14 × 103/μL on admission. Anticoagulation in septic thrombophlebitis, including classic Lemierre's syndrome, remains controversial and is not routinely recommended in the absence of thrombus propagation or clot in a critical location. On follow-up imaging, the hepatic vein thrombophlebitis remained stable without progression or new embolic events, and liver function normalized in parallel with antimicrobial therapy alone.

Empiric therapy was initiated with meropenem and doxycycline for broad-spectrum coverage while cultures were pending. Once blood cultures identified F. necrophorum, therapy was de-escalated on day 5 to piperacillin/tazobactam and metronidazole, a combination chosen on the basis of clinical judgement and in vitro susceptibility testing.

Conservative, antibiotic-only management was pursued for this small hepatic abscess, with percutaneous drainage deferred given the size of the lesion and the rapid clinical response. Limited evidence, drawn largely from case reports and small case series rather than controlled studies, suggests that antibiotic-only therapy can be effective for small pyogenic liver abscesses caused by Fusobacterium species when early clinical and radiographic improvement is observed, reserving drainage for larger collections or treatment failure [7, 8].

Most previously reported abdominal variants of Lemierre's syndrome involve septic thrombophlebitis of the portal or mesenteric venous system (pylephlebitis) rather than the hepatic vein itself; a recent systematic review identified 36 reported cases of the gastrointestinal variant of Lemierre's syndrome, and other literature reviews describe more than twenty cases of Fusobacterium-associated pylephlebitis [8]. True hepatic vein thrombophlebitis due to Fusobacterium is considerably rarer and has been described mainly in isolated reports [5]. The present case, with hepatic rather than portal or mesenteric vein thrombophlebitis, adds to this small body of literature and highlights that the visceral variant of Lemierre's syndrome extends beyond the portal venous system.

Clinicians evaluating young immunocompetent patients with polymetastatic septic presentations, severe cytopenias, and no oropharyngeal source should obtain cross-sectional chest and abdominal imaging and ensure that blood culture protocols include anaerobic identification technique and extended anaerobic incubation, since F. necrophorum grows slowly and may be missed by standard processing windows if anaerobic cultures are not specifically requested or incubated long enough; this consideration is one of the most actionable diagnostic points raised by this case. In patients with relevant sexual exposure history, anorectal mucosal disease should be actively sought as a candidate portal of entry, even when anorectal symptoms are absent.